How Execution Risk Can Change a Strategy’s Outcome
Execution risk is what a commitment meets after it can no longer be unmade. The commitment was made before the year began. In the same twelve months the largest account stopped paying, after eleven years of buying and at 171 days to pay against 110 for the rest. Both are published, neither causes the other, and a risk that happened was not therefore foreseeable.
What happens when a commitment and a bad year land in the same twelve months?
Start with the two facts and resist arranging them. Anjani Stationers Private Limited, an invented manufacturer, makes hard-bound registers and exercise books in a single city and sells them to schools and institutions it can list by name.
Fact one was settled before the year opened. At the start of its second year the business committed money it could not get back, taking a shareholding in Chitra Binding Works Private Limited, and its standing baseThe part of a year's running cost that sits at one total whether the works turns out a handful of registers or a great many. Once a year's worth of it is agreed, the amount stops answering to how trade goes. rose from Rs 49,60,000/- to Rs 74,00,000/-. The rise is Rs 24,40,000/-, or 49.19 per cent on the year before.
Fact two happened somewhere inside the year. The Sunrise Public School group, a customer of eleven years standing, took Rs 81,00,000/- of the year's Rs 2,70,00,000/- of revenue, or 30.00 per cent. The same group held Rs 38,00,000/- of the Rs 95,00,000/- of money still owedAmounts customers have been invoiced for and have not yet paid. It sits on a balance sheet as an entitlement rather than as cash in the bank, which is why an account can be large here and invisible elsewhere. across all thirty six accounts, or 40.00 per cent. The group settles its invoices in 171 days against 110 for the other thirty five. And during that year, it stopped paying.
Something has already happened to those two facts. Almost every reader joins them. A business takes on cost, its biggest buyer walks away, the year goes badly, and the sentence writes itself in the space between the second full stop and the third. Nothing published anywhere joins those two facts, and that absence is worth more than the story would have been.
Picture a household that agrees a second rent in April for a room it will need all year. In September the scooter it runs to work on breaks down. Both landed in the same twelve months, and the telling flows better that way, so a neighbour retelling it will put the two in one sentence with a joining word in the middle. The rent and the scooter never touched each other. The rent and the scooter shared a year. Sharing a year is a fact about a calendar rather than about either of them.
So what is execution risk actually asking?
Execution risk is what a commitment meets after it can no longer be unmade. Execution risk is not a mood, not a score and not a general worry about whether people are any good at their jobs. The question is about a particular window: the stretch of time in which the cost is fixed and the trade is not.
Two neighbouring questions get confused with it constantly, and both are settled elsewhere in these notes rather than here. The first is whether the decision was sound. A sound decision is judged against what the commitment was aimed at, rather than against the year it landed in. The second is whether the year was good. A good year is a description of an outcome. Neither one is execution risk. The decision and the year run on different clocks, and execution risk is the name for everything that arrives between the two.
Take the sharpest practical form of it. The last moment at which this cost base could have been chosen differently was before the year opened. After that, every single thing the twelve months produced, good or bad, met a standing base already fixed at Rs 74,00,000/-. A month of strong selling met it. A quiet month met it. An account that stopped paying met it. The base did not consult any of them. Not consulting is precisely what makes a standing cost a standing cost.
A lease shows the same thing. Once twelve months are signed on a second stall in the market, everything that happens in month three arrives at a rent that was agreed in month zero. A slow week does not reduce it. A festival week does not increase it. The signing was the decision; the eleven months after it are the execution.
When was the last moment this cost base could have been chosen differently?
What was committed, and how much of it can a reader actually see?
The commitment itself is narrated in full elsewhere in these notes, and only its shape is needed here. In one year the standing base climbed by Rs 24,40,000/-. The notes carrying that year name what the money went on in six words: on people, on space and on a binding operation the business bought into.
The accounting notes carry an estimated split of the same rise, and the label travels with the figures every time they move. Employee benefits rose Rs 6,00,000/-. The fixed part of other operating costs rose Rs 11,40,000/-, mostly a second warehouse taken during the year. Depreciation and amortisation rose Rs 7,00,000/- on the assets bought. The three limbs sum to Rs 24,40,000/- exactly, and the caption that carries them says: the split is an estimate, not a disclosure.
Read that caption twice. A figure and its label are one object, and separating them manufactures a disclosure nobody made. An estimate that says so is better evidence than a disclosure that is missing, and the words are what make it so.
The split matters here for a reason that has nothing to do with disclosure. The reason is time. Every one of those three limbs is a recurring annual cost, so the commitment did not simply happen once and finish; it kept happening for the whole of the year. Employees are paid every month. A warehouse charges rent every month. Depreciation runs whether anybody looks at it or not. A cost that keeps arriving is what makes a commitment unmakeable in the middle rather than merely regrettable in hindsight.
What is published about what happened, and how much of it is there?
An account of execution usually starts inventing at exactly this point, so the count is made out loud instead. Exactly two things are published about what happened during that year, and saying the number aloud is what stops a writer reaching for a third that was never published.
The first is the year's own arithmetic. Revenue rose from Rs 2,40,00,000/- to Rs 2,70,00,000/-, up Rs 30,00,000/- and 12.50 per cent. ContributionSales money less the costs that move up and down with the selling. It is the pool that every unchanging cost has to be paid out of, and it is not the same thing as profit. rose from Rs 1,02,60,000/- to Rs 1,15,50,000/-, up Rs 12,90,000/-. The standing base rose Rs 24,40,000/-. Operating profit fell from Rs 53,00,000/- to Rs 41,50,000/-, down Rs 11,50,000/-, and the operating margin moved from 22.08 per cent to 15.37 per cent, down 6.71 points.
The second is the largest account. The customer notes print both sides of every ratio so a reader can rebuild the arithmetic instead of trusting the answer: Rs 81,00,000/- against Rs 2,70,00,000/-, Rs 38,00,000/- against Rs 95,00,000/-, and Rs 57,00,000/- against Rs 1,89,00,000/-. One name at 30.00 per cent of revenue, thirty five names averaging Rs 5,40,000/- each at about 2.00 per cent apiece, and a ten point gap between that one name's share of revenue and its share of what is owed. The customer notes give the gap a name: dependenceWhat concentration turns into once a business would struggle to replace a name if it lost it. The full treatment, and the test that separates the two, is covered separately in these notes.. The difference between that name and plain concentrationHow much of a business's trade sits with a small number of its customers. A measure of shape only, and the procedure for reading it runs elsewhere in these notes. is worked out in seven steps, covered separately.
Two published things are the whole of it, and an account that needs more than two is about to invent something. Say plainly what is missing rather than skating past it. Nothing anywhere records what this business attempted during those twelve months. Nothing records what it abandoned. Nothing records what it knew, or when it knew it, or what anybody decided in response to anything. An execution judgement would need all four of those materials, and none of them exist in the record.
Collection daysAn average count of the days between a customer being invoiced and that customer's money landing in the bank. It describes what happened rather than what anybody arranged. are a measurement of how long money took to come in. Collection days are not a target that was set and not something anybody agreed to and missed. A marker set beside a measurement quietly turns it into a failure, and nobody published the target that marker would stand for.
How many things does this library publish about what happened during that year?
The rise in the standing base was Rs 24,40,000/- and the rise in contribution was Rs 12,90,000/-. What does the difference between them come to, and what does it show?
Do the two facts join up?
The join is the hardest step, and the answer turns out to be arithmetic rather than an appeal to careful reasoning. The tempting sentence, written out in full, runs: the largest account stopped paying, so the profit fell.
Now run the year’s own subtraction. One line settles the whole question. The rise in the standing base was Rs 24,40,000/-. The rise in contribution was Rs 12,90,000/-. Subtract the second from the first and Rs 11,50,000/- comes out. The published fall in operating profit is Rs 11,50,000/- exactly, to the rupee, with no residue left over for any other cause at all.
Read what that closure does. The closure is not a caution and not a plea for humility. There is no room in the published arithmetic for the account to have moved that line, and that is a much stronger statement than saying nobody has claimed it did. A reader who checks whether a subtraction leaves a residue never has to argue about whether a story is plausible.
Refusing everything would be as dishonest as claiming everything, so be precise about what the account did touch. An account that stops paying moves the money still owed and it moves how long money takes to come in. Money owed and collection days are not the same object as an operating result. The year’s published lines are a contribution and a standing base rather than a cash statement, so anything said about money owed carries that limit in the same breath.
Two bills arrive at a household in the same week. One can be pointed to in the bank statement, line by line, with the balance changing as it goes. The other is a promise somebody made that has not turned into money yet. Both bills feel the same when the envelopes are opened. The two live in completely different parts of the accounts.
An account that stops paying moves which published figures?
Could anybody outside have seen it coming?
Both easy answers are less interesting than what the customer notes actually say, so start with their own words. Both figures, the 30.00 per cent of revenue and the 40.00 per cent of the money owed, were sitting on the sheet a full year before the Sunrise Public School group stopped paying.
So a signal was readable, and that much is published rather than argued. Anybody holding those two sheets could see one name at 30.00 per cent while thirty five names sat at about 2.00 per cent apiece, and could see that name holding a larger share of what was owed than of what was sold.
Most readers slide at exactly this point, so draw the line precisely. A share of revenue and a share of the money owed describe a shape of exposure. The two shares name no date. The two shares name no outcome. Neither says which name will do anything, and thirty five other accounts sat on the same sheet with the same kind of figures beside them. A risk that happened is not evidence that it was foreseeable, and a signal that was readable is not the same object as an event that was predicted.
The reverse error is the one a careful reader makes, so give it equal weight. Refusing to notice a readable signal, on the grounds that hindsight makes everything look obvious, is its own mistake. The shape of exposure was genuinely there, in print, a year ahead. The honest position holds both halves at once: one of the two was published and the other was not.
A householder knows the roof is old, has known for years, and anybody who climbed up would agree. Knowing that is not the same as knowing which night it will leak, and the person who says the leak was foreseeable is quietly claiming to have held the second piece of knowledge when only the first was available.
The customer notes record both shares standing on the sheet a full twelve months ahead of the stopping. What follows?
What would a business have to publish before this question could be answered?
Turn the refusal into something a reader can use. Rather than saying the record is thin and stopping there, write down exactly what would be needed, item by item, with each item a real document rather than a wish. Six things.
| The document that would settle it | Why a reader would need it | Published here |
|---|---|---|
| The terms agreed with that account, and whether they had been changed | Because 171 days against 110 could be an arrangement or a drift, and the two read completely differently | nothing |
| How long each part of what was owed had been outstanding, rather than one total | Because Rs 38,00,000/- sitting for one month and the same amount sitting for a year are different objects | nothing |
| Whether any amount had been provided against or written off before, and when | Because an earlier write-off is a business already saying what it thinks of an account | nothing |
| That account's own accounts | Because the state of the buyer is the one thing that would actually speak to what happened | nothing |
| Whether the order bookWork already asked for and not yet delivered. It is a count of promises in hand rather than of money received, and the difference matters more than it looks. carried a commitment or only a habit | Because eleven years of buying and a signed order for next season are not the same evidence | nothing |
| What the same list of thirty six names had done in earlier years | Because one account stopping once is one observation, and one observation has no shape | nothing |
Anjani Stationers publishes none of the six. No filing anywhere asks a company to lodge any of the six, so leaving all six blank is what very nearly every business does, and it hides nothing whatever. An execution verdict written on material of that thinness measures how sure the writer felt, rather than how the business ran.
A list of what is missing, with a reason beside each line, is something a reader can hand on and argue with, and a judgement is not. Give somebody a score and their only two moves are agreement and disagreement. Give them six named papers and they are already halfway to the next request worth making. A note that opens a conversation differs from one that closes it in exactly that way.
Somebody asks how well this business executed, and the only material available is its published accounts. What can honestly be written down?
The write-up that explained the year, and every fact in it was published
An analyst is writing up the second year. Three things sit in front of them and all three are true. The business committed money it could not get back. Operating profit fell Rs 11,50,000/-. The largest account, thirty per cent of revenue and eleven years old, stopped paying.
The write-up says what any reader would expect it to say: the business took on cost and then its biggest customer stopped paying, so the year came in short. The sentence is fluent, every noun in it is sourced, and it is wrong.
Here is what actually went wrong, and the obvious diagnosis is the wrong one: not a single figure was invented and not a single one was misread. The invented part is the word so. Three published facts were arranged in the order a story wants them, and the joining words between them came from the writer rather than from the record.
The arithmetic refuses it outright. Outright refusal is what makes this failure refutable rather than merely suspect. The rise in the standing base less the rise in contribution is the fall in operating profit, to the rupee, with no residue left over for any other cause at all. There is no room in the published subtraction for that account to have moved that line, and an account that stops paying moves what is owed and how long money takes to come in rather than what a contribution less a standing base returns.
Now the cost, landed somewhere specific rather than described as a worse note. A reader downstream now carries a business with a customer problem. The accounts describe a different business. The next question asked is about the customer list rather than about the commitment. The commitment is never examined again. And the one thing the year actually settled, that a fixed base rose faster than the contribution carrying it, drops out of the file entirely.
The part worth sitting with is this. The story was not rejected for being unlikely. The story was rejected because the arithmetic had already closed without it, and a reader who checks for a residue never has to argue about plausibility at all.
The fix is one line, and it is not a more cautious tone: a closed subtraction has no gap to explain, so check whether the arithmetic already closes before reaching for anything to fill one.
The write-up said the business took on cost and then its biggest customer stopped paying, so the year came in short. Which part of that was invented?
So what may honestly be written down?
The output is shorter than the reader wants it to be. Four lines, in this order, each one a thing that could be pasted into a note and defended across a table.
One. The commitment and its size: a standing base climbing by Rs 24,40,000/- in the second year, spent on the three things the rivalry notes name in six words, with the three limbs estimated elsewhere at Rs 6,00,000/-, Rs 11,40,000/- and Rs 7,00,000/-, and that source's own words attached, that the split is an estimate, not a disclosure.
Two. The year's arithmetic and its closure: contribution rose Rs 12,90,000/-, the standing base rose Rs 24,40,000/-, operating profit fell Rs 11,50,000/-, and the subtraction closes with no residue.
Three. The account: Rs 81,00,000/- of the year's Rs 2,70,00,000/- of revenue, or 30.00 per cent; Rs 38,00,000/- of the Rs 95,00,000/- of money still owed, or 40.00 per cent; 171 days to collect where the rest take 110; eleven years of buying; and it stopped paying. Print both sides of each ratio, every time.
Four. The statement that nothing published joins the first three, together with the six named documents that would be needed before anybody could.
Four lines is the honest output and a paragraph about execution failure is not, and the four lines survive a challenge while the paragraph does not. Somebody can push back on any one of the four, and the source it came off can be handed over. A push back on the paragraph meets nothing underneath but the writer's own arrangement of things.
One act is one observation. A reader who has watched one commitment through one year has learned the shape of the question, and nothing whatever about how often the answer comes out either way.
How a reader tests an execution claim before writing it
Four lines travel with any execution claim, in this order, and they are worked here on this business rather than listed as advice.
One, when was the commitment made and when could it last have been chosen differently? Here, before the year opened. Everything after that moment met a base already fixed at Rs 74,00,000/-, and everything before that moment belongs to a different question entirely.
Two, what is published about what happened, named item by item, with the count said out loud? Here, two: the year's own arithmetic and the largest account. Saying the number aloud is what stops a writer quietly reaching for a third.
Three, does the arithmetic leave room for the story? Here, none at all: Rs 24,40,000/- less Rs 12,90,000/- is Rs 11,50,000/-, the published fall exactly, and the residue is zero.
Four, which papers would have to exist before the question could be settled? Here, six of them, written as documents rather than as wishes, and every one of them absent.
An execution claim with all four lines blank is a story rather than a finding, and line three alone would have stopped the write-up above before anybody read a word about the customer.
Which single check would have stopped that write-up before anybody argued about the customer?
Why can no control be built from what is published?
Every quantity that would make an interesting control here is one nobody published.
A control that slid the moment of the stopping backwards and forwards through the year would need a schedule, and no schedule exists anywhere. A control that attached some measure of how prone an account is to stopping would need exactly the advance record shown above to be absent. And a control that moved the money still owed and redrew the two shares would be arithmetically sound and still somebody else’s subject: those two shares and the gap between them belong to the customer notes.
In its place stands the one subtraction, printed in full and drawn as a movement with its residue shown as an empty box. The subtraction can be checked with a pencil. A pencil check is a stronger form of interaction than a control, and it survives printing.
The largest account settles in 171 days against 110 for the other thirty five. What kind of figure is that?
What India supplies here, and what it does not
India supplies the currency, the lakh and crore digit grouping, the legal form Private Limited, a school year that fills an order book in spring, and the existence of a regime under which companies place accounts on public record. The regime publishes totals, and it asks for none of the six items in the card above. The card stands empty for that reason rather than concealed.
The mechanism itself carries no border at all. A commitment that cannot be unmade in the middle of a year is the same object in every jurisdiction on earth, and a published account that records what came in rather than what was attempted is the ordinary shape of a filing everywhere. The mechanism rests on no rate, no threshold and no filing requirement, so a change in any of the three would leave it exactly as it was. What a disclosure regime asks a company to lodge does move from year to year, and the register itself is the only place that answers it.
The window, the record and the four lines. A commitment that can no longer be unmade meets a year, the published record of that year holds two facts and no line joining them, and four lines are what a reader may honestly write down at the end.
Narrating what the business committed to and why is covered separately under Corporate and Business Strategy Compared: Where and How to Win. Building cases on the business’s own published lines is covered separately under How to Build Business Scenarios for a Company. The customer list procedure, sorting a list of names, and working out what a share of revenue becomes when one buyer holds it are all covered separately under How to Analyse Customer Concentration and Dependence. The effect of the two routes to growth on a reported growth rate is covered separately under Organic and Inorganic Growth Compared: One Year, Two Rates. The claim of the people who put the money in on what the year left is covered separately under Growth Investment vs Capital Return: One Pot, Two Uses. Sorting published material into facts, inferences and conditionals is covered separately under How to Separate Facts, Inference and Scenarios in Company Research. How a purchase of shares is arranged, priced or reviewed afterwards is covered separately in the material on transactions and corporate finance.
Where the one named source stands, and what the arithmetic above rests on
| Source | What it is | Treatment in this guide | Where |
|---|---|---|---|
| Ministry of Corporate Affairs | The register where companies place their annual accounts on public record | This row stands behind one sentence and no more: that a rule exists somewhere obliging a company to lodge its accounts where anybody can read them. Not a rule, not a limit and not a number out of that regime is quoted above. The six lines left blank in the card are blank because lodging accounts has never meant lodging those. | mca.gov.in |
| The arithmetic in this guide | The working behind the one subtraction shown above | Every amount above belongs to invented parties and to no trading ones. The year's lines and the account's lines are quoted from material already built in these notes rather than worked out afresh here, and the estimated split of the standing base rise carries its source's label wherever it appears. No filing, no survey and no trade study stands behind a single amount above. | finmaverick.com |
Anjani Stationers Private Limited, Chitra Binding Works Private Limited and the Sunrise Public School group are invented.
Educational material. Not advice on any investment, tax, budget or market position.